- Email: [email protected]
Alveolar hypoxia, inhibition of hypoxic pulmonary vasoconstriction, and permeability edema
Respiration Physiology, 75 (1989) 11-18
11
Elsevier RSP 01487
Alveolar hypoxia, inhibition of hypoxic pulmonary vasoconstriction, and permeability edema Frederick W. Cheney, Barbara L. Eisenstein and Michael J. Bishop Departments of Anesthesiology and Medicine, University of Washington, School of Medicine, Seattle, Washington, U.S.A. (Accepted for publication 8 October 1988) Abstract. We previously reported that regional alveolar hypoxia reduces oleic acid-induced permeability edema formation [Cheney etal. (1987). J. Appl. Physiol. 62: 1690-1697]. In order to define the role of hypoxic pulmonary vasoconstriction (HPV) on this effect, we studied the effects of regional alveolar hypoxia on permeability edema formation with this response inhibited. Dogs weighing 25 + ! kg in which the HPV response had been inhibited by the administration ofminoxidil(i mg/kg i.v.)were anesthetized, mechanically ventilated and had a bronchial divider placed so the lef~ lower lobe (LLL) could be ventilated with an Flo2 = 0.05 or Flo2 --- 1, while the right lung was continuously ventilated with an Flo2 = 1.0. In 10 study animals the LLL was ventilated with an FIo, = 0.05 for 4 h after induction of bilateral permeability pulmonary edema with 0.05 ml/kg of intravenous oleic acid. In six control animals the LLL was ventilated with an F l o 2 - - 1 for 4 h after the same injury. Postmortem gravimetric analysis indicates that alveolar hypoxia of the LLL with the HPV response inhibited had no effect on pulmonary edema formation. We conclude that inhibition of HPV abolishes the protective effect of regional alveolar hypoxia on oleic acid-induced permeability edema formation.
Dog; Edema; Flow diversion; Hypoxia; Olcic acid; Pulmonary circulation
We have observed that regional alveolar hypoxia reduces oleic acid-induced permeability pulmonary edema formation (Cbeney et al., 1987). In a study of dogs with oleic acid lung injury, we found that ventilation of the left lower lobe (LLL) with a hypoxic gas mixture significantly reduced edema formation as compared with the contralateral lung which was ventilated with 100% 02. We had initially hypothesized that the mechanism for this fmding was a reduction in microvascular pressure in the capillary bed distal to the precapillary site of hypoxic pulmonary vasoconstriction (Dawson, 1984). However, we found that HPV did not engender a decrease in pulmonary capillary pressure as measured by the double occlusion technique in the in vivo LLL (Cheney et al., 1987). We then hypothesized that alveolar hypoxia exerted an effect on pulmonary vascular permeability independent of the HPV response. In the present study, we Correspondence address: Frederick W. Cheney, Department of Anesthesiology, University of Washington, School of Medicine RN-10, Seattle, WA 98195, U.S.A. 0034-5687/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
12
F. CHENEY et al.
inhibited the hypoxic pulmonary vasoconstrictor response with the vasodilator minoxidil prior to inducing permeability pulmonary edema with oleic acid. We found that the effect of alveolar hypoxia on permeability edema was dependent on the HPV response in that its inhibition abolished the protective effect we had observed previously.
Methods
Instrumentation. Twenty mongrel dogs of either sex weighing 25 _+ 1 kg (SEM) were treated 16 h prior to the experiment with 1 mg/kg intravenous minoxidil diluted in propylene glycol and H20. Minoxidil is a smooth muscle dilator which has been shown to block hypoxic pulmonary vasoconstriction (Bishop et al., 1985) and has a duration of action of 48-72 h (Gottlieb et al., 1972; Lowenthal and Affrime, 1980; Bishop and Cheney, 1983). Animals were anesthetized with pentobarbital sodium (30 mg/kg i.v.), paralyzed with 100 mg of succinylcholine intramuscularly and their tracheas were intubated. The tidal volume was 15 ml/kg with an FIo2 = 1.0 and the respiratory rate was adjusted to maintain Paco2 between 30 and 35 mm Hg. Pentobarbital and succinylcholine were repeated as necessary to maintain anesthesia and paralysis. Arterial and (Swan-Ganz) pulmonary arterial catheters were placed via cutdown. Through a left thoracotomy the left upper lobe was resected and an electromagnetic flow transducer (In Vivo Metric Systems, Healdsburg, CA) was placed around the left main pulmonary artery. A right thoracotomy and isolation and loose banding of the right pulmonary artery was done in all animals to control for possible effects of thoracotomy and manipulation of the lungs on gravimetric comparisons between the left and right lower lobes of each animal. The surgical procedure was completed within 2 h. Both thoracotomies were loosely covered with plastic to conserve heat and moisture, the animal was placed supine and a bronchial divider placed through a tracheostomy. The tidal volume was apportioned between the sides by adjusting the Harvard double piston respirator such that the airway pressures (monitored with a Statham PM 131 PC transducer) were equal to the pre-lobectomy value on each side. The lungs were kept on 5 cm H20 PEEP to compensate for the absence of distending transpulmonary pressure with the chest open. Arterial (BP), pulmonary arterial (PPA)and pulmonary arterial wedge (Pw) pressures were monitored with Statham P 231D pressure transducers. All pressure and blood flow outputs (Zepeda Flow Meter, Seattle, WA) were recorded continuously on a Grass Model 7 polygraph. Cardiac output (0T) was obtained by the thermodilution method using an Edwards 9520A cardiac output computer. Protocol. Approximately 1 h after completion of the surgical procedure, Be, PPA, Pw, arterial and mixed venous blood gases, LLL blood flow (QLLL) and (~T were recorded with the LLL FIo2 set at 1.0 in both groups. Hypoxic challenge with hypoxic gas mixture ventilation of the LLL was carded out in all animals to assure that the HPV response had been blocked by the minoxidil. The hypoxic gas mixture consisted of 92% N2-5 %
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA
13
02-3 % CO2. After a stabilization period of 20 min the hemodynamic and blood gas measurements were repeated while the LLL was ventilated with the hypoxic gas mixture. The LLL FIo2 was returned to 1.0 in both groups and oleic acid (0.05 ml/kg body weight) was injected over 2 rain into the right atrium. In the control group (n = 6) the LLL FIo2 remained at 1.0, while in the study group (n = 10) the LLL FIo2 was changed to 0.05 immediately post-injury. Measurements were made hourly for 4 h post-injury in both groups. To evaluate the effects of probe placement or anatomy on the amount of right vs left oleic acid injury, the above protocol was repeated in four additional animals with the flow probe placed on the right rather than the left main pulmonary artery. In these animals the FXo~ was 1.0 bilaterally. At 4 h all animals were killed with pentobarbital overdose, the RLL and LLL were rapidly removed, weighed, and prepared for gravimetric analysis. Blood-free wet weightto-dry weight (W/D)ratios, estimates of total lung water and total blood-free dry weight were made using the cyanomethemoglobin method to estimate blood content (Pearce et al., 1965). Statistical comparisons of data between the 2 groups of animals were done using the unpaired t-test. Gravimetric data from the RLL and LLL of the same animal were compared with paired t-statistics, p values of < 0.05 were considered to represent significantdifferences. All results presented are mean values + standard error of the mean. Data analysis.
Results There was no significant difference in lobar weight, bloodless lobar water, or bloodless lobar dry weight between LLL and RLL in the study (LLL FIo2 = 0.05) or control (LLL TABLE 1 Mean values + SEM for gravimetric data in study and control animals. There were no statistical differences between study and control values for any variable. Study animals LLL Fro2 = 0.05 (n = 10)
Control animals LLL Flo2 = i.0 (n = 6)
Lobar weight (g)
L R
172 + 10 180+8
153 + 15 157+!4
Bloodless lobar
L R
115 + 4 120 + 5
103 + 10 108 + 10
Bloodless lobar dry weight (g)
L R
14A + 0.8 14.2 + 07
13.7 + 1.1 13.6 + 1.0
Wet weight/dry weight (g/g)
L R
H20 (g)
* p < 0.05 paired t.
8.92 + 0.26 9.43 + 0.31"
8A5 + 0.23 8.95 _+ 0.27*
14
F. CHENEY et al. TABLE 2
Mean values + SEM for gravimetric data in four additional control animals (LLL Fro2 = 1.0) who had the flow probe placed on the right pulmonary artery. There are no statistical differences between RLL and LLL (paired t). Control animals LLL F l o 2 -- i.0 (n ffi 4) Lobar weight (g)
L R
160 + 9 149 + 12
Bloodless lobar HzO (g)
L R
111 _+ 7 102 _+ 11
Bloodless lobar dry weight (g)
L R
13.5 +_ 0.80 12.6 __ 0.97
Wet weight/dry weight (g/g)
L R
9.24 + 0.44 9.12 + 0.39
FIo2 = 1.0) animals (table 1). The mean blood-free W/D ratio for the LLL was significantly lower than the RLL in both study and control animals (table 1). There were no differences between any of the LLL or RLL gravimetric variables of the study compared to the control animals (unpaired t) (table 1). In the four additional control animals in which the flow probe was placed on the right pulmonary artery there was no difference in gravimetric indices including W/D ratio between LLL and RLL (table 2). Prior to lung injury the percent of cardiac output perfusing the left lower lobe (0LLL/0T) decreased slightly when the LLL of study and control animals was ventilated with the hypoxic gas mixture (table 3). The decrease in 0LLL/0T from 27 + 1 to 25 + 2 during hypoxia of the LLL was significant in the study animals prior to injury (table 3). There was no change from LLL hyperoxia for values of PPA, Pw, (~LLL o r PVRLLL during LLL hypoxia in either group (table 3). There were significant increases in (~T, Paco2 and decreases in pH and PVo2 in the study animals during LLL hypoxia (table 3). The only difference between study and control groups (unpaired t) prior to injury was a 3 mm Hg higher Paco 2in study as compared to control animals during LLL hypoxia. After oleic acid injury the data for the cardiorespiratory variables are presented as the mean of the hourly measurements. The only differences in these variables between the study and control animals (unpaired t) were the Pao2 and a 3 mm Hg higher Paco2 in the study group (table 3). (~LLL a n d QLLL/(~Twere slightly lower in the study animals but the differences were not significant (table 3).
Discussion The small changes in (~LLL/0T in response to alveolar hypoxia of the LLL in both control and study animals prior to lung injury indicate that minoxidil pretreatment
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA
15
TABLE 3 Mean values + SEM for hemodynamic and blood gas data in study (LLL F[o2 = 0.05) and control (LLL F]o2 = l) animals before and after oleic acid lung injury. Pre-injury values Bilateral hyperoxia
PPA (mm Hg)
Study Control
Pw (mm Hg)
Study Control
t)T (L/rain)
Study Control
3.26 _+ 0.33 3.85 4- 0.57
3.56 4. 0.36* 3.95 + 0.38
2.93 + 0.38 3.02 + 0.37
0LLL
Study Control
860 _ 85 982 _+ 146
878 + 95 1009 _+ 121
689 _+ 88 849 + 133
Study Control
27 4. 2 26 _+ I
25 + 2* 25 _+ 1
24 + 2 27 __ 2
PVRLLL (mm Hg)/ (L/min)
Study Control
12 4. 2 i I 4. 1
14 _+ 2 10 + 0
20 _+ 2 16 4. 2
Pao2 (mm Hg)
Study Control
555 _+ 12 498 +_ 29
95 4. 9* 102 _+ 14'
66 + 3 393 + 27*
Paco 2 (mm Hg)
Study Control
33 _+ 1 31 _ 1
36 + 1' 33 4. !*
33 4. 0 30 _+ !*
pH
Study Control
7.41 _+ 0.03 7.42 4. 0.01
7.38 + 0.03* 7.40 4- 0.01
7.39 4. 0.02 7.40 4- 0.02
Pvoz (mm Hg)
Study Control
57 4- 3 57 4. 6
47 4. 2* 44 _+ 3
39 4. 2 50 _+ 5
(ml/min)
0LLL/QT x ! O0
16_ 1 18 _ 2
LLL hypoxia
7 _ 0.6 7 _+ 0.4
17 4- 1 17 4- 2
Average post-injury Study -- LLL hypoxia Cont. = bilateral hyperoxia
6 _+ 0.3 7 + 0.3
21 _. 1 20 + 2 8 4- 0.5 7 + 0.4
*p < 0.05 unpaired t between study and control groups. *p < 0.05 paired t c.f. bilateral hyperoxia pre-injury.
significantly blunted the HPV response. The small difference in (~LLL/(~T is significant in the study group probably only because there are more animals in that group (n = 10) than in the control group (n = 6) (table 3). In any event, the reduction in (~LLL/(~T from 27 + 2 to 25 + 2 is not likely to be physiologically significant. The most relevant point is that after oleic acid injury there is no difference in (~LLL/QT and (~LLL between study and control animals (table 3), which indicates that there is little if any flow diversion from the hypoxic LLL of the study animals. This is in contrast to the results in our
16
F. CHENEY et al. HPV IHTACT
t
30-
I
250 0t m
HPV INHIBITEO
I
,
20
I
X 15 I-
.0 10.,J ..J ._1
"0
5
0 Conlrol
Study
Group
Group
Control Group
Study Group
Fig. 1. Mean values + SEM for left lower lobe blood flow as a percent ofthe total cardiac output (t)LLL/t)T) during the post-injury period in control (LLL FIo2 = 1) and study animals (LLL Flo2 = 0.05) with the HPV response intact and with the response inhibited with minoxidil. The HPV intact data are from a previous study in our laboratory (Cheney et aL, 1987) and the HPV inhibited data are from the pres-nt study, t", P < 0.05 (unpaired t control vs study).
previous study (Cheney et al., 1987) with an identical protocol in which the HPV response was intact and there was a significant difference in t)LLL/0T between study and control animals (fig. 1) after oleic acid injury. After injury 0LLL was much higher in the minoxidil-treated animals during hypoxia (689 + 88 ml) than in the animals with the HPV response intact (295 + 35 ml) (Cheney et al., 1987). This is due to the higher t~T engendered by the minoxidil itself. Minoxidil dilates the systemic as well as the pulmonary vasculature so there is a reflex increase in t)T. The gravimetric results in the present study with the HPV response inhibited are also in contrast to our earlier study, in which this response was intact (Cheney et aL, 1987). In that study (Cheney et al., 1987) the hypoxie left lower lobes had a mean bloodless lobar water which was significantly lower than the fight lower lobes, which were hyperoxic throughout the study (fig. 2). In the present study with HPV inhibited there was no difference in bloodless lobar water between the hypoxic LLL and hyperoxic RLL (fig. 2). In the previous study with HPV intact (Cheney et aL, 1987), values for LLL and RLL bloodless dry weight were 15 + 1 and 17 + I g (P < 0.05), respectively, wlfile in the present study with HPV inhibited, there were no differences between hypoxic and hyperoxic lobes (table 1). With the HPV response intact in the previous study (Cheney et al., 1987) the W/D ratio for hypoxic LLL was 8.1 _+0.31 and the hyperoxic LLL W/D ratio was 8.95 + 0.32 (P < 0.05) while there were no differences in the W/D ratios for the LLL and RLL in control animals (Fio~ = 1.0 bilaterally). In the present study, with HPV inhibited, however, the LLL W / D ratios were significantly lower than the RLL W/D ratios in both study and control animals (table 1).
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA HPV INTACT
HPV INHIBITED
HPV INTACT
150". 125.
HPV INHIBITED
+I-
+
++
100-
S w
17
75-
U,I l-q[
0 Z
"~ .J
25o
L
R
L
R
STUDY GROUPS
i
III
L
R
L
R
CONTROL GROUPS
Fig. 2. Mean vaLaes _+ SEM for bloodless lung water content of left lower lobes (L) and right lower lobes (R) in experimental animals with the HPV response intact and another group with the response inhibited with minoxidil. The HPV intact data are from a previous study in our laboratory (Cheney et al., 1987) and the HPV inhibited data are from the present study. I-I, Lobar Fno2 = !.0; k~, lobar Flo2 = 0.05; *, P < 0.05 (paired t).
We think the reason for this difference in W/D ratio between LLL and RLL in both the study and control animals with HPV inhibited is due to the presence of the flow probe on the left main pulmonary artery during the injection of oleic acid. Minoxidil caused significant dilatation of the pulmonary arteries, so that even our largest flow probes caused some restriction of flow. As the flow probe was in place on the left pulmonary artery during the right atrial injection of oleic acid, it is possible that the LLL received slightly less oleic acid than the right lung. This is supported by the results in the 4 control animals in which the flow meter was placed on the right pulmonary artery. In these animals the LLL W/D ratio was slightly higher than the RLL, although the difference did not reach statistical significance. Because there was a similar difference in W/D ratio between the two lobes ventilated with 100% oxygen in control animals (table 1), it would indicate that the difference in W/D ratios seen in the hypoxic LLL and hyperoxic RLL in the study animals is not related to alveolar hypoxia. The results of the present study may shed light on the mechanism by which alveolar hypoxia decreases permeability edema formation. In our previous study we postulated three different mechanisms for this effect: (1) hypoxia-induced reduction in pulmonary capillary pressure; (2) direct effect of alveolar hypoxia on permeability edema; and (3) a decrease in microvascular area for leakage (Cheney et al., 1987). In that study we showed that pulmonary capillary pressure is not reduced in the presence of regional alveolar hypoxia as measured in the LLL by the double occlusion technique, so the first mechanism is not likely to be operative. The results of the present study essentially rule
18
F. CHENEY etal.
out alveolar hypoxia itself as having any direct effect on permeability edema formation in the model, as there is no difference in bloodless lung water or dry weight between hypoxic LLL and hyperoxic RLL in the same animals (table 1). Our findings suggest that alveolar hypoxia with HPV intact may exert its effects on this model of permeability edema by reducing the area for microvascular leakage. Dawson et ai. (1983) have shown in normal pump-perfused dog LLL that HPV modifies pulmonary perfusion such that there is a cessation of flow in some parallel vascular pathways. Ifhypoxia-induced vasoconstriction stopped the flow in some leaking parallel channels, then edema formation could be expected to decrease. Inhibition of the HPV response with minoxidil would therefore have allowed continued perfusion of the leaking vessels during LLL alveolar hypoxia. While the results of this study do not specifically prove this hypothesis, they do indicate that when diversion of flow away from an hypoxic lung lobe is blunted, edema formation is no longer affected by regional alveolar hypoxia. It should be noted that the results are specific only for the model studied in which the size of the hypoxic compartment (LLL) allows flow diversion to hyperoxic right lung without development of significant pulmonary hypertension. If the hypoxic compartment were large such that there was an increase in pulmonary artery pressure, then minoxidil might reduce edema formation if it produced a reduction in PPA. In conclusion, in the open chest canine oleic acid model of diffuse lung injury, inhibiton of the HPV response with the smooth muscle vasodilator minoxidil abolishes the protective effects of regional alveolar hypoxia on permeability edema formation. The most likely mechanism is that minoxidil prevents hypoxia-induced closure of leaking parallel vascular pathways in the lung. Acknowledgements.The authors thank Holly M. Kabinofffor typing and editing the manuscript. This work was supported by NIH Grants no. 24163 and no. 30542.
References Bishop, M.J. and F.W. Cheney (1983). Comparison of the effects of minoxidil and nifedipine on hypoxic pulmonary vasoconstriction in dogs. J. Cardiovasc. Pharmacol. 5: 184-189. Bishop, M.J., T. Huang and F.W. Cheney (1985). Effect of vasodilator treatment on the resolution of oleic acid injury in dogs. Am. Rev. Respir. Dis. 131: 421-425. Cheney, F.W., M.J. Bishop, E.Y. Chi and B.L. Eisenstein (1987). The effect of regional alveolar hypoxia on permeability pulmonary edema formation in dogs. J. Appl. Physiol. 62: 1690-1697. Dawson, C.A., T.A. Bronikowski, J.H. Linehan and T.S. Hakim (1983). Influence of pulmonary vasoconstriction on lung water and perfusion heterogeneity. J. Appl. Physiol. 54: 654-660. Dawson, C.A. (1984). Role of pulmonary vasomotion in physiology of the lung. Physiol. Rev. 64: 544-616. Gottlieb, T.B., R.C. Thomas and C.A. Chidsey (1972). Pharmacokinetic studies of minoxidil. Clin. Pharmacol. Ther. 13: 436-441. Lowenthal, D.T. and M. B. Affrime (1980). Pharmacology and pharmacokinetics ofminoxidil. J. Cardiovasc. Pharmacol. 2 (Suppl.): 93-106. Pearce, M.L., J. Yamashita and J. Beazell (1965). Measurement of pulmonary edema. Circ. Res. 16: 482-488.
11
Elsevier RSP 01487
Alveolar hypoxia, inhibition of hypoxic pulmonary vasoconstriction, and permeability edema Frederick W. Cheney, Barbara L. Eisenstein and Michael J. Bishop Departments of Anesthesiology and Medicine, University of Washington, School of Medicine, Seattle, Washington, U.S.A. (Accepted for publication 8 October 1988) Abstract. We previously reported that regional alveolar hypoxia reduces oleic acid-induced permeability edema formation [Cheney etal. (1987). J. Appl. Physiol. 62: 1690-1697]. In order to define the role of hypoxic pulmonary vasoconstriction (HPV) on this effect, we studied the effects of regional alveolar hypoxia on permeability edema formation with this response inhibited. Dogs weighing 25 + ! kg in which the HPV response had been inhibited by the administration ofminoxidil(i mg/kg i.v.)were anesthetized, mechanically ventilated and had a bronchial divider placed so the lef~ lower lobe (LLL) could be ventilated with an Flo2 = 0.05 or Flo2 --- 1, while the right lung was continuously ventilated with an Flo2 = 1.0. In 10 study animals the LLL was ventilated with an FIo, = 0.05 for 4 h after induction of bilateral permeability pulmonary edema with 0.05 ml/kg of intravenous oleic acid. In six control animals the LLL was ventilated with an F l o 2 - - 1 for 4 h after the same injury. Postmortem gravimetric analysis indicates that alveolar hypoxia of the LLL with the HPV response inhibited had no effect on pulmonary edema formation. We conclude that inhibition of HPV abolishes the protective effect of regional alveolar hypoxia on oleic acid-induced permeability edema formation.
Dog; Edema; Flow diversion; Hypoxia; Olcic acid; Pulmonary circulation
We have observed that regional alveolar hypoxia reduces oleic acid-induced permeability pulmonary edema formation (Cbeney et al., 1987). In a study of dogs with oleic acid lung injury, we found that ventilation of the left lower lobe (LLL) with a hypoxic gas mixture significantly reduced edema formation as compared with the contralateral lung which was ventilated with 100% 02. We had initially hypothesized that the mechanism for this fmding was a reduction in microvascular pressure in the capillary bed distal to the precapillary site of hypoxic pulmonary vasoconstriction (Dawson, 1984). However, we found that HPV did not engender a decrease in pulmonary capillary pressure as measured by the double occlusion technique in the in vivo LLL (Cheney et al., 1987). We then hypothesized that alveolar hypoxia exerted an effect on pulmonary vascular permeability independent of the HPV response. In the present study, we Correspondence address: Frederick W. Cheney, Department of Anesthesiology, University of Washington, School of Medicine RN-10, Seattle, WA 98195, U.S.A. 0034-5687/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
12
F. CHENEY et al.
inhibited the hypoxic pulmonary vasoconstrictor response with the vasodilator minoxidil prior to inducing permeability pulmonary edema with oleic acid. We found that the effect of alveolar hypoxia on permeability edema was dependent on the HPV response in that its inhibition abolished the protective effect we had observed previously.
Methods
Instrumentation. Twenty mongrel dogs of either sex weighing 25 _+ 1 kg (SEM) were treated 16 h prior to the experiment with 1 mg/kg intravenous minoxidil diluted in propylene glycol and H20. Minoxidil is a smooth muscle dilator which has been shown to block hypoxic pulmonary vasoconstriction (Bishop et al., 1985) and has a duration of action of 48-72 h (Gottlieb et al., 1972; Lowenthal and Affrime, 1980; Bishop and Cheney, 1983). Animals were anesthetized with pentobarbital sodium (30 mg/kg i.v.), paralyzed with 100 mg of succinylcholine intramuscularly and their tracheas were intubated. The tidal volume was 15 ml/kg with an FIo2 = 1.0 and the respiratory rate was adjusted to maintain Paco2 between 30 and 35 mm Hg. Pentobarbital and succinylcholine were repeated as necessary to maintain anesthesia and paralysis. Arterial and (Swan-Ganz) pulmonary arterial catheters were placed via cutdown. Through a left thoracotomy the left upper lobe was resected and an electromagnetic flow transducer (In Vivo Metric Systems, Healdsburg, CA) was placed around the left main pulmonary artery. A right thoracotomy and isolation and loose banding of the right pulmonary artery was done in all animals to control for possible effects of thoracotomy and manipulation of the lungs on gravimetric comparisons between the left and right lower lobes of each animal. The surgical procedure was completed within 2 h. Both thoracotomies were loosely covered with plastic to conserve heat and moisture, the animal was placed supine and a bronchial divider placed through a tracheostomy. The tidal volume was apportioned between the sides by adjusting the Harvard double piston respirator such that the airway pressures (monitored with a Statham PM 131 PC transducer) were equal to the pre-lobectomy value on each side. The lungs were kept on 5 cm H20 PEEP to compensate for the absence of distending transpulmonary pressure with the chest open. Arterial (BP), pulmonary arterial (PPA)and pulmonary arterial wedge (Pw) pressures were monitored with Statham P 231D pressure transducers. All pressure and blood flow outputs (Zepeda Flow Meter, Seattle, WA) were recorded continuously on a Grass Model 7 polygraph. Cardiac output (0T) was obtained by the thermodilution method using an Edwards 9520A cardiac output computer. Protocol. Approximately 1 h after completion of the surgical procedure, Be, PPA, Pw, arterial and mixed venous blood gases, LLL blood flow (QLLL) and (~T were recorded with the LLL FIo2 set at 1.0 in both groups. Hypoxic challenge with hypoxic gas mixture ventilation of the LLL was carded out in all animals to assure that the HPV response had been blocked by the minoxidil. The hypoxic gas mixture consisted of 92% N2-5 %
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA
13
02-3 % CO2. After a stabilization period of 20 min the hemodynamic and blood gas measurements were repeated while the LLL was ventilated with the hypoxic gas mixture. The LLL FIo2 was returned to 1.0 in both groups and oleic acid (0.05 ml/kg body weight) was injected over 2 rain into the right atrium. In the control group (n = 6) the LLL FIo2 remained at 1.0, while in the study group (n = 10) the LLL FIo2 was changed to 0.05 immediately post-injury. Measurements were made hourly for 4 h post-injury in both groups. To evaluate the effects of probe placement or anatomy on the amount of right vs left oleic acid injury, the above protocol was repeated in four additional animals with the flow probe placed on the right rather than the left main pulmonary artery. In these animals the FXo~ was 1.0 bilaterally. At 4 h all animals were killed with pentobarbital overdose, the RLL and LLL were rapidly removed, weighed, and prepared for gravimetric analysis. Blood-free wet weightto-dry weight (W/D)ratios, estimates of total lung water and total blood-free dry weight were made using the cyanomethemoglobin method to estimate blood content (Pearce et al., 1965). Statistical comparisons of data between the 2 groups of animals were done using the unpaired t-test. Gravimetric data from the RLL and LLL of the same animal were compared with paired t-statistics, p values of < 0.05 were considered to represent significantdifferences. All results presented are mean values + standard error of the mean. Data analysis.
Results There was no significant difference in lobar weight, bloodless lobar water, or bloodless lobar dry weight between LLL and RLL in the study (LLL FIo2 = 0.05) or control (LLL TABLE 1 Mean values + SEM for gravimetric data in study and control animals. There were no statistical differences between study and control values for any variable. Study animals LLL Fro2 = 0.05 (n = 10)
Control animals LLL Flo2 = i.0 (n = 6)
Lobar weight (g)
L R
172 + 10 180+8
153 + 15 157+!4
Bloodless lobar
L R
115 + 4 120 + 5
103 + 10 108 + 10
Bloodless lobar dry weight (g)
L R
14A + 0.8 14.2 + 07
13.7 + 1.1 13.6 + 1.0
Wet weight/dry weight (g/g)
L R
H20 (g)
* p < 0.05 paired t.
8.92 + 0.26 9.43 + 0.31"
8A5 + 0.23 8.95 _+ 0.27*
14
F. CHENEY et al. TABLE 2
Mean values + SEM for gravimetric data in four additional control animals (LLL Fro2 = 1.0) who had the flow probe placed on the right pulmonary artery. There are no statistical differences between RLL and LLL (paired t). Control animals LLL F l o 2 -- i.0 (n ffi 4) Lobar weight (g)
L R
160 + 9 149 + 12
Bloodless lobar HzO (g)
L R
111 _+ 7 102 _+ 11
Bloodless lobar dry weight (g)
L R
13.5 +_ 0.80 12.6 __ 0.97
Wet weight/dry weight (g/g)
L R
9.24 + 0.44 9.12 + 0.39
FIo2 = 1.0) animals (table 1). The mean blood-free W/D ratio for the LLL was significantly lower than the RLL in both study and control animals (table 1). There were no differences between any of the LLL or RLL gravimetric variables of the study compared to the control animals (unpaired t) (table 1). In the four additional control animals in which the flow probe was placed on the right pulmonary artery there was no difference in gravimetric indices including W/D ratio between LLL and RLL (table 2). Prior to lung injury the percent of cardiac output perfusing the left lower lobe (0LLL/0T) decreased slightly when the LLL of study and control animals was ventilated with the hypoxic gas mixture (table 3). The decrease in 0LLL/0T from 27 + 1 to 25 + 2 during hypoxia of the LLL was significant in the study animals prior to injury (table 3). There was no change from LLL hyperoxia for values of PPA, Pw, (~LLL o r PVRLLL during LLL hypoxia in either group (table 3). There were significant increases in (~T, Paco2 and decreases in pH and PVo2 in the study animals during LLL hypoxia (table 3). The only difference between study and control groups (unpaired t) prior to injury was a 3 mm Hg higher Paco 2in study as compared to control animals during LLL hypoxia. After oleic acid injury the data for the cardiorespiratory variables are presented as the mean of the hourly measurements. The only differences in these variables between the study and control animals (unpaired t) were the Pao2 and a 3 mm Hg higher Paco2 in the study group (table 3). (~LLL a n d QLLL/(~Twere slightly lower in the study animals but the differences were not significant (table 3).
Discussion The small changes in (~LLL/0T in response to alveolar hypoxia of the LLL in both control and study animals prior to lung injury indicate that minoxidil pretreatment
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA
15
TABLE 3 Mean values + SEM for hemodynamic and blood gas data in study (LLL F[o2 = 0.05) and control (LLL F]o2 = l) animals before and after oleic acid lung injury. Pre-injury values Bilateral hyperoxia
PPA (mm Hg)
Study Control
Pw (mm Hg)
Study Control
t)T (L/rain)
Study Control
3.26 _+ 0.33 3.85 4- 0.57
3.56 4. 0.36* 3.95 + 0.38
2.93 + 0.38 3.02 + 0.37
0LLL
Study Control
860 _ 85 982 _+ 146
878 + 95 1009 _+ 121
689 _+ 88 849 + 133
Study Control
27 4. 2 26 _+ I
25 + 2* 25 _+ 1
24 + 2 27 __ 2
PVRLLL (mm Hg)/ (L/min)
Study Control
12 4. 2 i I 4. 1
14 _+ 2 10 + 0
20 _+ 2 16 4. 2
Pao2 (mm Hg)
Study Control
555 _+ 12 498 +_ 29
95 4. 9* 102 _+ 14'
66 + 3 393 + 27*
Paco 2 (mm Hg)
Study Control
33 _+ 1 31 _ 1
36 + 1' 33 4. !*
33 4. 0 30 _+ !*
pH
Study Control
7.41 _+ 0.03 7.42 4. 0.01
7.38 + 0.03* 7.40 4- 0.01
7.39 4. 0.02 7.40 4- 0.02
Pvoz (mm Hg)
Study Control
57 4- 3 57 4. 6
47 4. 2* 44 _+ 3
39 4. 2 50 _+ 5
(ml/min)
0LLL/QT x ! O0
16_ 1 18 _ 2
LLL hypoxia
7 _ 0.6 7 _+ 0.4
17 4- 1 17 4- 2
Average post-injury Study -- LLL hypoxia Cont. = bilateral hyperoxia
6 _+ 0.3 7 + 0.3
21 _. 1 20 + 2 8 4- 0.5 7 + 0.4
*p < 0.05 unpaired t between study and control groups. *p < 0.05 paired t c.f. bilateral hyperoxia pre-injury.
significantly blunted the HPV response. The small difference in (~LLL/(~T is significant in the study group probably only because there are more animals in that group (n = 10) than in the control group (n = 6) (table 3). In any event, the reduction in (~LLL/(~T from 27 + 2 to 25 + 2 is not likely to be physiologically significant. The most relevant point is that after oleic acid injury there is no difference in (~LLL/QT and (~LLL between study and control animals (table 3), which indicates that there is little if any flow diversion from the hypoxic LLL of the study animals. This is in contrast to the results in our
16
F. CHENEY et al. HPV IHTACT
t
30-
I
250 0t m
HPV INHIBITEO
I
,
20
I
X 15 I-
.0 10.,J ..J ._1
"0
5
0 Conlrol
Study
Group
Group
Control Group
Study Group
Fig. 1. Mean values + SEM for left lower lobe blood flow as a percent ofthe total cardiac output (t)LLL/t)T) during the post-injury period in control (LLL FIo2 = 1) and study animals (LLL Flo2 = 0.05) with the HPV response intact and with the response inhibited with minoxidil. The HPV intact data are from a previous study in our laboratory (Cheney et aL, 1987) and the HPV inhibited data are from the pres-nt study, t", P < 0.05 (unpaired t control vs study).
previous study (Cheney et al., 1987) with an identical protocol in which the HPV response was intact and there was a significant difference in t)LLL/0T between study and control animals (fig. 1) after oleic acid injury. After injury 0LLL was much higher in the minoxidil-treated animals during hypoxia (689 + 88 ml) than in the animals with the HPV response intact (295 + 35 ml) (Cheney et al., 1987). This is due to the higher t~T engendered by the minoxidil itself. Minoxidil dilates the systemic as well as the pulmonary vasculature so there is a reflex increase in t)T. The gravimetric results in the present study with the HPV response inhibited are also in contrast to our earlier study, in which this response was intact (Cheney et aL, 1987). In that study (Cheney et al., 1987) the hypoxie left lower lobes had a mean bloodless lobar water which was significantly lower than the fight lower lobes, which were hyperoxic throughout the study (fig. 2). In the present study with HPV inhibited there was no difference in bloodless lobar water between the hypoxic LLL and hyperoxic RLL (fig. 2). In the previous study with HPV intact (Cheney et aL, 1987), values for LLL and RLL bloodless dry weight were 15 + 1 and 17 + I g (P < 0.05), respectively, wlfile in the present study with HPV inhibited, there were no differences between hypoxic and hyperoxic lobes (table 1). With the HPV response intact in the previous study (Cheney et al., 1987) the W/D ratio for hypoxic LLL was 8.1 _+0.31 and the hyperoxic LLL W/D ratio was 8.95 + 0.32 (P < 0.05) while there were no differences in the W/D ratios for the LLL and RLL in control animals (Fio~ = 1.0 bilaterally). In the present study, with HPV inhibited, however, the LLL W / D ratios were significantly lower than the RLL W/D ratios in both study and control animals (table 1).
HYPOXIC PULMONARY VASOCONSTRICTION AND PERMEABILITY EDEMA HPV INTACT
HPV INHIBITED
HPV INTACT
150". 125.
HPV INHIBITED
+I-
+
++
100-
S w
17
75-
U,I l-q[
0 Z
"~ .J
25o
L
R
L
R
STUDY GROUPS
i
III
L
R
L
R
CONTROL GROUPS
Fig. 2. Mean vaLaes _+ SEM for bloodless lung water content of left lower lobes (L) and right lower lobes (R) in experimental animals with the HPV response intact and another group with the response inhibited with minoxidil. The HPV intact data are from a previous study in our laboratory (Cheney et al., 1987) and the HPV inhibited data are from the present study. I-I, Lobar Fno2 = !.0; k~, lobar Flo2 = 0.05; *, P < 0.05 (paired t).
We think the reason for this difference in W/D ratio between LLL and RLL in both the study and control animals with HPV inhibited is due to the presence of the flow probe on the left main pulmonary artery during the injection of oleic acid. Minoxidil caused significant dilatation of the pulmonary arteries, so that even our largest flow probes caused some restriction of flow. As the flow probe was in place on the left pulmonary artery during the right atrial injection of oleic acid, it is possible that the LLL received slightly less oleic acid than the right lung. This is supported by the results in the 4 control animals in which the flow meter was placed on the right pulmonary artery. In these animals the LLL W/D ratio was slightly higher than the RLL, although the difference did not reach statistical significance. Because there was a similar difference in W/D ratio between the two lobes ventilated with 100% oxygen in control animals (table 1), it would indicate that the difference in W/D ratios seen in the hypoxic LLL and hyperoxic RLL in the study animals is not related to alveolar hypoxia. The results of the present study may shed light on the mechanism by which alveolar hypoxia decreases permeability edema formation. In our previous study we postulated three different mechanisms for this effect: (1) hypoxia-induced reduction in pulmonary capillary pressure; (2) direct effect of alveolar hypoxia on permeability edema; and (3) a decrease in microvascular area for leakage (Cheney et al., 1987). In that study we showed that pulmonary capillary pressure is not reduced in the presence of regional alveolar hypoxia as measured in the LLL by the double occlusion technique, so the first mechanism is not likely to be operative. The results of the present study essentially rule
18
F. CHENEY etal.
out alveolar hypoxia itself as having any direct effect on permeability edema formation in the model, as there is no difference in bloodless lung water or dry weight between hypoxic LLL and hyperoxic RLL in the same animals (table 1). Our findings suggest that alveolar hypoxia with HPV intact may exert its effects on this model of permeability edema by reducing the area for microvascular leakage. Dawson et ai. (1983) have shown in normal pump-perfused dog LLL that HPV modifies pulmonary perfusion such that there is a cessation of flow in some parallel vascular pathways. Ifhypoxia-induced vasoconstriction stopped the flow in some leaking parallel channels, then edema formation could be expected to decrease. Inhibition of the HPV response with minoxidil would therefore have allowed continued perfusion of the leaking vessels during LLL alveolar hypoxia. While the results of this study do not specifically prove this hypothesis, they do indicate that when diversion of flow away from an hypoxic lung lobe is blunted, edema formation is no longer affected by regional alveolar hypoxia. It should be noted that the results are specific only for the model studied in which the size of the hypoxic compartment (LLL) allows flow diversion to hyperoxic right lung without development of significant pulmonary hypertension. If the hypoxic compartment were large such that there was an increase in pulmonary artery pressure, then minoxidil might reduce edema formation if it produced a reduction in PPA. In conclusion, in the open chest canine oleic acid model of diffuse lung injury, inhibiton of the HPV response with the smooth muscle vasodilator minoxidil abolishes the protective effects of regional alveolar hypoxia on permeability edema formation. The most likely mechanism is that minoxidil prevents hypoxia-induced closure of leaking parallel vascular pathways in the lung. Acknowledgements.The authors thank Holly M. Kabinofffor typing and editing the manuscript. This work was supported by NIH Grants no. 24163 and no. 30542.
References Bishop, M.J. and F.W. Cheney (1983). Comparison of the effects of minoxidil and nifedipine on hypoxic pulmonary vasoconstriction in dogs. J. Cardiovasc. Pharmacol. 5: 184-189. Bishop, M.J., T. Huang and F.W. Cheney (1985). Effect of vasodilator treatment on the resolution of oleic acid injury in dogs. Am. Rev. Respir. Dis. 131: 421-425. Cheney, F.W., M.J. Bishop, E.Y. Chi and B.L. Eisenstein (1987). The effect of regional alveolar hypoxia on permeability pulmonary edema formation in dogs. J. Appl. Physiol. 62: 1690-1697. Dawson, C.A., T.A. Bronikowski, J.H. Linehan and T.S. Hakim (1983). Influence of pulmonary vasoconstriction on lung water and perfusion heterogeneity. J. Appl. Physiol. 54: 654-660. Dawson, C.A. (1984). Role of pulmonary vasomotion in physiology of the lung. Physiol. Rev. 64: 544-616. Gottlieb, T.B., R.C. Thomas and C.A. Chidsey (1972). Pharmacokinetic studies of minoxidil. Clin. Pharmacol. Ther. 13: 436-441. Lowenthal, D.T. and M. B. Affrime (1980). Pharmacology and pharmacokinetics ofminoxidil. J. Cardiovasc. Pharmacol. 2 (Suppl.): 93-106. Pearce, M.L., J. Yamashita and J. Beazell (1965). Measurement of pulmonary edema. Circ. Res. 16: 482-488.